Clinical pathology of plastic ingestion in marine birds and relationships with blood chemistry 1 2

Pollution of the environment with plastic debris is a significant and rapidly expanding threat to biodiversity due to its abundance, durability, and persistence. Current knowledge of the negative effects of debris on wildlife is largely based on consequences that are readily observed, such as entanglement or starvation. Many interactions with debris, however, result in less visible and poorly documented sub-lethal effects, and as a consequence, the true impact of plastic is underestimated. We investigated the sub-lethal effects of ingested plastic in Flesh-footed Shearwaters (Ardenna carneipes) using blood chemistry parameters as a measure of bird health. The presence of plastic had a significant negative effect on bird morphometrics and blood calcium levels, and a positive relationship with the concentration of uric acid, cholesterol, and amylase. That we found blood chemistry parameters to be related to plastic pollution is one of few examples to date of the sublethal effects of marine debris and highlights that superficially healthy individuals may still experience the negative consequences of ingesting plastic debris. Moving beyond crude measures, such as reduced body mass, to physiological parameters will provide much needed insight into the nuanced and less visible effects of plastic.


31
Increasing demand for, and production of, plastic products coupled with inadequate waste 32 management and policy contributes to the ongoing and rapidly expanding issue of pollution of our 33 waterways and wildlife 1, 2 . Mass production of small plastic items, such as pellets or 'nurdles' combined 34 with the break-up of larger items present in the marine environment for decades has led to an increase 35 in the availability of bite-size pieces for wildlife 3, 4 and a concomitant increase in the frequency of 36 accidental consumption by marine animals in all of the world's oceans 5 .

37
The negative consequences resulting from interactions between wildlife and plastic debris are 38 diverse, often visually striking, and can include nutritional deprivation 6 , entanglement 7,8 , and damage to 39 or obstruction of the gut (e.g., perforations and ulcers; 9 ). Many of these interactions also include less 40 visible and therefore less well documented effects such as reduced growth and survival rates following 41 ingestion 6, 10, 11 and as a consequence, we are drastically underestimating the true impact of plastic 42 waste on our oceans.

43
Plastic is inherently toxic 12 and becomes increasingly hazardous over time as it accumulates 44 pollutants from the surrounding marine environment 13,14 . Once ingested, the absorbed toxins leach into 45 the animal's blood stream 6,11,15,16 and contribute to neurological, behavioural, and reproductive 46 problems at all levels of biological organization, and in extreme cases, death of individuals 11,17,18 . At 47 least 43-100% of the world's marine mammal, seabird, and turtle species are at risk from the ingestion of 48 plastic 19 and this number is expected to increase as plastic production and studies into its effects also 49 rise 20 . How we manage our waste and understand the true scope and severity of impacts this has on 50 wildlife therefore has broad implications for the health of marine ecosystems.

51
The Flesh-footed Shearwater (Ardenna carneipes) is one of the most heavily impacted marine 52 vertebrates with regards to plastic pollution 6 , containing a mixture of macro-, micro-, and ultrafine 53 debris 21 . Populations across the species' range are in decline [22][23][24] with the ingestion of plastic implicated 54 in the downward trend due to its negative effect on chick growth and survival and its capacity to act as a 55 vector for contaminants 6,25 . The mechanisms by which plastic negatively influences shearwater and 56 other wildlife populations are poorly understood 18 , are likely present at the molecular or cellular levels of organization, and may not result in organisms' death, but in their poor health 18,26

68
Jackson 27 , and detailed in Lavers,et al. 28 . In brief, seawater (approx. 150 ml) at ambient temperature 69 was gently pumped into the proventriculus through a tube, thus displacing any ingested items. Once fluid 70 and stomach contents began to flow back up the oesophagus (i.e., once the stomach was filled 71 completely), the bird was inverted over a container to collect anything expelled. Plastic items were dried 72 and weighed to the nearest 0.001 g using an electronic balance. 73 74 2.2. Blood sampling. Approximately 0.5 ml of whole blood was collected from the brachial vein of each 75 bird using a 26-gauge needle and stored in lithium heparin vacutainer tubes (Grenier MiniCollect, 76 Austria). A drop of blood was smeared onto a microscope slide immediately after collection and air dried.

77
Samples were refrigerated at 4°C overnight then express posted to IDEXX Laboratory (Rydalmere, NSW,

78
Australia by airplane and courier the following afternoon (i.e., within 18 hours of collection). At the 79 laboratory, two slides were prepared from each sample using two separate techniques to evaluate 80 haematological parameters. The first technique involved diluting and mixing 20 μl of whole blood with 81 620 μl of Eosin stain solution. The mixture was then used to fill two chambers of a haemocytometer for 82 manual count of heterophils and eosinophils through a microscope. The second technique required a 83 May-Grunswald/Giemsa stain to perform a differential stain, calculate a total white blood cell count, and assess morphology of the red blood cells, white blood cells, and platelets. The proportion of different 85 types of leucocytes (lymphocytes, heterophils, monocytes, eosinophils, and basophils) was assessed 86 based on an examination of 100 leucocytes under oil immersion (1000× magnification). Blood 87 components (glucose (mmol/L)), urea (mmol/L), calcium (mmol/L), total protein (g/L), albumin (g/L),  log-transformed to improve normality. Parameters were grouped by physiological function: those 99 associated with the liver (aspartate aminotransferase, glutamate dehydrogenase, bile acids), blood 100 proteins (total protein, albumen, globulin), those with kidney/renal function (urea, uric acid), and white 101 blood cell types (numbers of heterophils, lymphocytes, monocytes, eosinophils, and basophils). Other 102 parameters (glucose, calcium, cholesterol, creatine kinase, amylase) and the number of white blood cell 103 types were analysed individually.

104
Relationships between measurements of body condition or blood parameters and plastic 105 ingestion were explored using a series of univariate (reporting F-statistics) or multivariate linear models 106 (reporting Wilk's λ, followed by univariate tests when results were significant). Analyses were done using 107 R 3.5.1 (R Core Team 2018), differences were considered statistically significant when p < 0.05 and values 108 are reported as mean ± SD.

RESULTS
We found no relationship between the presence, number, or mass of plastics and packed cell volume, 112 white blood cell count, or white blood cell composition (Table 2). Similarly, there was no relationship 113 between any measure of plastic burden and molecules associated with liver function (aspartate 114 aminotransferase, glutamate dehydrogenase, or bile acids; Table 2).

115
The only significant relationship between plastic burden and measures of kidney function was a 116 positive association between the number of plastic pieces and concentration of uric acid ( Figure 1, Table   117 2). There were negative relationships between plastic presence, number, and mass with blood calcium, 118 and positive relationships between plastic presence and cholesterol, as well as plastic mass and amylase 119 concentration (Table 2).

120
The presence of plastic had significant negative relationships with all morphometrics, though 121 there was no link between bird size and the number or mass of ingested plastic ( Figure 1, Table 2).

124
There is scant information on blood composition for free-living birds, especially seabirds, many of which 125 have been identified as threatened species. Flesh-footed Shearwaters have experienced considerable 126 mortality through human actions (e.g., fisheries bycatch) which contributed to a reduction in population 127 numbers in recent decades 24, 31 . Our data therefore provide much needed blood chemistry reference 128 values for Flesh-footed Shearwaters, a species that was recently up-listed to Near Threatened 30 and is 129 declining across its range [22][23][24] . Blood chemistry data also provides valuable information on other threats 130 that may affect the health of individuals, including plastic debris.

131
Blood cell volume (PCV) varied considerably among individual Flesh-footed Shearwaters and is 132 known to be influenced by age or physiological status in other Procellariforms 32, 33 . The body mass 133 recorded for some shearwater fledglings (180-270 g) was well below the range typically observed in this 134 species (500-750 g) 34 , suggesting these birds were in poor condition. This agrees with our visual 135 observations, which indicated these individuals were emaciated and often lethargic. However, PCV 136 values do not reliably reflect condition in some seabird species, even when birds are experimentally 137 handicapped 35,36 . In light of this, and the variability in blood cell count parameters we observed, it is unsurprising we failed to detect a relationship between PCV and plastic presence, mass, or number of 139 ingested items (Table 2).

140
Changes in blood parameters have been evaluated in some seabirds well adapted to seasonal 141 fasting (e.g., penguins; 37 ). Fasting in birds occurs in three phases with the final phase indicating the 142 depletion of fat reserves, when body proteins are catabolized, resulting in a higher rate of body-mass 143 loss, which is often indicative of a critical limit 38 . Low total protein (<25 g/L) is considered a good 144 indicator of starvation in birds, but can also signify chronic disease or stress 39 . In Flesh-footed 145 Shearwaters, total protein was only slightly above this threshold (33 ± 10 g/L; Table 1) and was not 146 related to plastic presence, mass, or number of ingested items (Table 2) (Table 2), which is attributed to poor fat reserves and increased chemical exposure 6 .

176
In contrast to previous work 6 , we did not find significant relationships between the amount of 177 plastic and bird morphometrics. Rather, simply the presence of plastic had a negative relationship with 178 body mass, wing length, culmen, and head+bill length. Though our sample size is small (n = 38), it does 179 span the range of severity of plastic ingestion 6, 52 . This suggests that such effects are likely to be complex, 180 influenced by inter-annual variation, and could imply a threshold over which additional plastic has little 181 additional effect. Understanding this mechanism is fundamental to the question "how much plastic is 182 bad", which is key for establishing relevant policy targets.

183
To date, studies of the effect of plastic debris on wildlife have been largely confined to 184 quantifying plastic burden (e.g., Provencher, et al. 53 ), or relating that burden to gross measures of body 185 condition, such as mass or chick size (e.g., Lavers, Bond and Hutton 6 ). Though these studies remain 186 important, increasingly, the potential for sublethal impacts on birds' physiology is of concern 54 , including 187 the presence of toxic chemicals 6, 15, 55 , harmful microbiota 56, 57 , and the ubiquity of small particles which 188 can be distributed throughout the digestive tract 21, 58 . Such relationships have been recently described in 189 marine invertebrates and fish 59, 60 , but few data exist for seabirds. This is concerning as birds can 190 compensate for physiological impairments caused by disease or other stressors 49 meaning they can appear healthy for a period of time, which can be misleading when undertaking visual assessments of 192 health and condition or using superficial measures, such as body size.

193
While quantifying population-level impacts of plastic ingestion on wildlife remains an important 194 goal 18, 61 , this will continue to be challenging given the myriad threats faced by seabirds 62 , and as we